Figure 2: Western Blot of His-tagged C. tetani antigen with anti His-tag HRP Conjugate. The soluble fraction of the overexpressed protein was used for this Western Blot. The expected molecular weight is 50 kDa. Anti- His Tag HRP conjugated antibody was used in a 1:1000 dilution in blocking buffer.

The antigen used for detection of anti-tetanus antibodies is derived from the tetanus neurotoxin and commonly used for in vitro testing. For preparation of the DiaCHIP, this antigen had to be overexpressed in E. coli and purified from whole cell lysate. To verify the successful purification of the antigen, Western Blot analysis was performed (figure 2). The usage of an antibody against the N-terminal His-tag revealed a protein with a molecular weight of about 50 kDa. This correlates with the expected molecular weight of the antigen ¹⁾.

Figure 3: ELISA for the tetanus antigen with blood samples taken after the immunization. The graph shows the absorption over time for TMB. The red line shows the binding of anti-tetanus antibodies out of the blood serum to purified tetanus antigen and the blue line shows the binding of HRP-Streptavidin to bBSA. The grey lines show the corresponding negative controls.

We performed an ELISA to investigate, whether the blood samples we obtained contain antibodies that bind specific to our purified tetanus antigen. Therefore, we spotted tetanus antigen and bBSA of the same concentration (12.5 µg/mL) and 2 times PBS as negative controls. The tetanus antigen and one PBS well were incubated with 5 mg/mL BSA, blood serum, and HRP-labeled anti-human for 1 h, respectively. The bBSA and the other PBS well were incubated with HRP-labeled Streptavidin (0.5 µg/mL) for 1  h (positive control). After each step the wells were washed with PBS. To visualize the binding, Tetramethylbenzidine was added and the absorption at 650 nm was measured in a plate reader over time (figure 3). The absorption signal for tetanus is even higher than the positive control, which illustrates the successful binding of antibodies from the blood to tetanus antigen.

Figure 4: Binding curves for the detection of the binding of anti Tetanus antibodies from blood serum via iRIf.

After the validation of the antigen antibody binding for tetanus by ELISA we wanted to see if we can detect this interaction also with iRIf. Therefore we first expressed our Clostridium tetani antigen and GFP in E.coli. The purified His-tagged proteins then were spotted on a PDITC surface. The slide with the immobilized proteins was flushed with human blood serum, and anti human antibodies. Strong binding was detected for the tetanus antigen spots, while the GFP spots only exhibited minimal binding (see figure 1A-D). This indicates that components contained in the blood serum bind specifically to the tetanus antigen. The subsequent binding of anti human antibodies to the tetanus antigen spots (figure 2) confirms that the proteins that bound during the blood serum step are antibodies. We therefore successfully demonstrated that it is possible to detect the binding of antibodies from blood serum to antigens, we immobilized on a surface.

Figure 5: Quotion picture of iRIf measurement. Line A was flushed with serum taken before vaccination against tetanus. Line B was flushed with serum taken 2 weeks after vaccination. Spots from left to right show: GFP, neg. control, Tetanus antigen.

With an additional experiment we wanted to see if we, in principle, can differentiate the vaccination status of patients with our system. Therefore we used blood serum samples, taken before and two weeks after a tetanus boost vaccination. A slide with immobilized tetanus antigen and GFP was prepared (see figure 3). It was consecutively flushed with both blood sera. The individual binding of the two sera is seen in figure 3B and 3C. The blood serum taken after vaccination exhibits a much more intense binding to the tetanus spot than the serum taken before binding. This observation is supported by the binding curves obtained by this measurements (see figure 3D). The validation that the exhibited binding, shown by the protein spots during the exposure to blood serum is due to antibodies could not be performed, due to problems with the anti human antibody supply.

Figure 6:Quotion pictures of slides flushed with blood serum taken before and after vaccination. (A) Schematic pattern of on glass slide immobilized proteins. (B) Quotient picture of iRIf measurement in which the slide was flushed with anti GFP and blood serum taken previous to vaccination . (C) Quotient picture of iRIf measurement in which the slide was flushed with anti GFP and blood serum five weeks after vaccination.

While the outcome of this first experiment looked promising for our future application to determine the vaccination status of patients, further experimentation did not confirm these initial results. We were still able to detect binding of both blood sera to the tetanus antigen, but no difference in binding intensity was visible (See figure 4B and C). To test the assumption that the used blood was not fresh enough and therefore lost some of its binding capabilities a fresh blood sample five weeks after vaccination was taken for the experiment shown in figure 4. The increase in relative light intensity at distinct spots over the course of the experiment is visualized in figure 5 and can be correlated with the amount of antibodies binding to the respective antigen spot. This indicates that the DiaCHIP enables quantification of antibody titers in addition to detecting their presence.

We could successfully show that the cell-free expression of GFP-His results in functional GFP proteins that are bound by anti-GFP antibodies during an iRIf measurent. The DNA coding for GFP fused to a 10x His-tag was added to the DiaMIX and expression was performed for 2 h. The DiaMIX containing cell-free expressed GFP-His was then spotted on an iRIf slide with a Ni-NTA surface by hand (figure 1). The spots were incubated on the slide over night.

Figure 1: Specific Ni-NTA surface on iRIf glass slide.

This specific surface allows binding of the expressed GFP-His via the His-tag while all the other proteins present in the DiaMIX are not bound to the surface. This step was followed by the blocking and washing protocol to prevent further unspecific binding. The iRIf slide was then flushed with a anti-GFP antibody solution, to analyze the binding to the cell-free expressed GFP-His. As a positive control we spotted a purified GFP-His onto the surface as well as DiaMIX that did not contain any DNA as negative control (figure 2).
Find out more about the preparation of the DiaCHIP by producing a protein microarray from a DNA template using cell-free expression.

The measurement shows a high signal for the cell-free expressed GFP-His and the positive control as can also be seen in the respective binding curve shown in figure 3.

Figure 2: Measurement of cell-free expressed GFP on Ni-NTA surface. The presence of GFP is detected with GFP antibodies within an iRIf measurement The spot on the right is the cell-free expressed GFP-His protein spotted onto the Ni-NTA surface. The spot on the left shows the positive control, GFP-His expressed in E. coli, purified and spotted onto the Ni-NTA surface. The negative control represents the DiaMIX without DNA. A clear iRIf signal is detected for the positive control and the cellfree expressed GFP.

Figure 3: Binding curve of anti-GFP binding to cell-free expressed GFP-His on a Ni-NTA surface. Cell-free expressed GFP-His was spotted onto a specific Ni-NTA surface and flushed with anti-GFP. The binding curve indicated a binding event at the respective spot.

▲ Close Box

To test our device we aimed at reproducing measurements that we were previously able to perform in our commercial device. For this reason we focused on using two antibodies: anti-GFP and polyclonal anti-rabbit. GFP and rabbit derived anti-HCV (Hepatitis C Virus) antibodies were used as antigens in this experiment. Note that the anti-HCV antibodies only served as binding partners for the anti-rabbit antibodies, since no HCV proteins were used in this experiment. The antigens were spotted onto an iRIf slide whose binding layer consisted of an APTES/PDITC surface. The spots on the slide were produced by pipetting 3.5 µl [1 mg/ml] rabbit-anti-HCV protein and 3.5 µl [1 mg/ml] GFP onto the slide and incubated overnight. After incubation the slide was blocked for 30 min in BSA solution. A Canon 50D camera was used to record the measurement and was set to take one picture every 5 seconds. The exposure time was set in order for the pixels in the image to be approx. 80% of maximum light saturation before the solution was flushed onto the chip. The antibody solutions were pipetted into the flow-chamber without the use of any microfluidic device. Instead, a syringe was loaded with 660 µl [5 µg/ml] anti-GFP antibody solution (diluted in PBS) and connected to the input pipe of our device. The content of the syringe was then slowly released via this pipe into the chamber of the device by gently dispensing the solution from the syringe. When the whole volume ran over the chip, the process was repeated with 660 µl [5 µg/ml] anti-rabbit antibody solution in the same way. The injection of both solutions was performed during approximately 45 minutes.

Figure 3: Quotient picture of the same measurement with favourable light conditions.

The quotient pictures shows all intensity shifts that occured during the measurement. Two left spots: GFP spots at which an intensity shift occured during anti-GFP flush. Right spot: anti-HCV antibody (derived from rabbit), at which an intensity shift occured during the flush with anti-Rabbit.

Figure 4: Comparison of two quotient pictures of our DiaCHIP with the professional setup. Left: Professional device; Right: Our device. Both quotient picture show the binding events that occured during an iRIf measurement with anti-GFP at the GFP spots.

Video 1 shows the results of the measurement. Both, binding of anti-GFP and anti-rabbit to the corresponding antigen spots, could be observed. Due to the fact that the experiment was performed on our demonstration device which was built with a transparent casing, fluctuations in surrounding light had a strong, detrimental influence on the measurement quality. To minimize the influence of the unstable surrounding light, the resulting pictures had to be averaged over 10 pictures each. This in turn lead to a more stable light situation, however the signal strengh dropped as a consequence. Figure 3 shows a quotient picture of the measurement where the light situation was temporarily appropriate. The problem of surrounding light scattering into the device can of course be overcome using a non-transparent casing. An evaluation of how well our device performs in comparison with the professional device, was not performed, mainly due to time constraints. However our results hint that we get comparable results for strong binding such as GFP/anti-GFP, as can be seen in figure 4.

Figure 2: Cell-free expressed GFP-His spotted on a PDITC slide. Cell-free expression was performed for 2 hours at 37°C. The positive control was purified GFP-His, the negative control was a sample of cell-free expression mix not containing DNA.

For cell-free protein expression lots of different proteins, like RNA-polymerases, ribosomes and other E. coli proteins, are essential. So during the process of cell-free expression all these proteins are, besides the target protein, also present in the microfluidic chamber. To have a sufficient amount of target protein immobilized on the chip we established a specific surface chemistry on the glass slide. After testing several tag systems we established a Ni-NTA surface because it worked best for us. Using a Ni-NTA covered glass slide (figure 1) we could increase the amount of bound cell-free expressed GFP-His compared to an unspecific surface (figure 2). In both experiments 1.5 µg of self purified GFP-His was used as positive control. The cell-free expression mix without DNA is supposed to result in no target protein and served as negative control. The cell-free reactions were performed for 2 hours at 37°C. The samples were pipetted by hand onto the surfaces and incubated on the slide overnight. In the iRIf device, the optical detection method we use, the slides were blocked with BSA and then flushed with anti-GFP antibodies. An increase in light intensity where the spotted protein was located represents a binding event. Due to less unspecific binding of proteins to the surface, the interaction of anti-GFP antibodies with cell-free expressed GFP-His is higher on the Ni-NTA surface.

Our DiaMIX is a complex system consisting of several individually adjusted components that allow us to make use of the cells own production system, without the tedious drawbacks of normal cell-based protein expression. As a main component of the DiaMIX the E. coli lysate contains the necessary transcription ans translation apperatus. The lysate is complemented by additives needed for energy regeneration, buffering and generation of the protein. An overview of cell-free expression systems and its components can be found on our project pages.
As ultimate evaluation, we compared the DiaMIX to a commercial kit as seen in figure 1.

Both expression systems successfully expressed the applied vector.
Furthermore, the cell-free expression of GFP with our cell-free mix could be verified in an iRIf measurement.
To obtain these final results, a lot of optimization and testing had to be done. Want to know more about how we established the DiaMIX? You can find detailed information about the most important experiments on the cell-free expression results page.